Experimental Evaluation of Suitability of Selected Multi-Criteria Decision-Making Methods for Large-Scale Agent-Based...

Area of Science:

• Computational Economics
• Decision Science

Background:

• Multi-criteria decision-making (MCDM) methods are crucial for complex system modeling.
• Agent-based computational economic (ACE) models require efficient decision-making algorithms for large-scale simulations.

Purpose of the Study:

• To compare the suitability of four MCDM methods (WPM, TOPSIS, VIKOR, PROMETHEE) for large-scale ACE models.
• To identify the most computationally efficient and effective MCDM method for simulations exceeding 10,000 agents.

Main Methods:

• Selected four MCDM methods: Weighted Product Model (WPM), Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS), VIKOR, and Preference Ranking Organization Method for Enrichment Evaluations (PROMETHEE).
• Conducted 12,800 tests across four hardware configurations, analyzing performance with and without server parameters.
• Utilized an illustrative decision-making scenario for comparative analysis.

Main Results:

• All tested MCDM methods are practically applicable.
• The VIKOR method yielded the best performance in the conducted tests.
• Computational efficiency varied across methods and hardware configurations.

Conclusions:

• The VIKOR method is recommended as the most suitable MCDM approach for large-scale ACE model simulations.
• The findings provide valuable insights for selecting appropriate decision-making tools in complex agent-based modeling.